WHAT IS IT?
We recently acquired a new Google Vision Kit which uses artificial intelligence to detect and recognize objects. Along with Google Voice Kit which was released late last year, these new kits blend making with artificial intelligence. Vision Kit includes an onboard Raspberry Pi camera connected to a Raspberry Pi Zero. It takes about 1.5 hours to assemble the components and another hour to install and configure the software and apps required. Below I'm "recognizing" a bottle of dog vitamins. The Vision Kit can connect to an LCD so that the image is displayed. When setting up the Kit, you pair the device to your Android phone so that it can run over your wireless network. There are more than 1,000 objects in the image library.
EXECUTING THE SCRIPT
Once everything is installed, you run a Python script using a Chrome browser shell command. Using the device IP address, you execute a pre-built script called 'Image_Classification_Camera.py' that automatically recognizes whatever is in the viewfinder. The command line scrolls constantly while is recognizes and classifies the object. I captured the image below from my laptop. Note that the script constantly updates as the camera detects whatever is in its viewfinder
SEE IT IN ACTION
Here's a promotional video illustrating how it works and illustrating Face Tracker, Face Recognition, and Image Classification.
No comments:
Post a Comment